# "GO FAST" Footage from Tom DeLonge's To The Stars Academy. Bird? Balloon?

Excellent posts by @Todd Wasson. I've been struggling to articulate the same since I share his flight simming experience but lack the math background to argue the points.

There are a a few critical and fundamental mistakes made in the assumptions that go into the models that is very frustrating to see from someone who is somewhat familiar with military aircraft systems being discussed here.

Indeed the trajectory of the plane has to be accounted to retrieve meaningful results. Considering a straight trajectory for the plane will put you way off. The fact that the plane and GoFast are not at the same altitude is important too, so the calculations are more accurate when building a 3D model : https://www.geogebra.org/m/axynssmq

My model gives a speed of 58 Knots for GoFast, considering the lines of bearing and RNG given in the video. Of course we don't know the wind shear between the plane altitude (25000ft) and GoFast (12000ft), that's an important factor that is missing to retrieve the GoFast speed.
This guys did the same kind of analyses and also find a speed of 50-60 Knots :

Source: https://youtu.be/y5Uf4N-JkQY

I've checked the video, thanks. Interesting results. This looks almost identical to Mick's bottom picture from earlier. I've checked the VC (just using their 2D diagram to see if it was in the ballpark):

Their UFO speed is 52 knots with flight path along vector (-0.64, 0.76) (measured from pixels). Looking for a VC of about 210 knots or so.

(Coordinate system: X+ right, Y+ up (altitude direction), Z+ forward (top of screen)

I use x and z here and omit y because this slide is 2D.

JetToUFODirection.x =-0.64
JetToUFODirection.z = 0.76

From the diagram the velocity vector of the UFO looks approximately the same as the long green diagonal line, so I'll just set them equal for now as I don't feel like writing code right now. Just trying to check if VC is in the ballpark here, not be super accurate with the numbers:

UFO.FlightDirection.x = JetToUFODirection.x
UFO.FlightDirection.z = JetToUFODirection.z

UFO.WorldFrame.Velocity = UFO.Speed * UFO.Direction
UFO.WorldFrame.Velocity.x = 52 (knots) *-0.64 (x direction vector= -33.28 knots
UFO.WorldFrame.Velocity.z = 52 (knots) * 0.76 = 39.52 knots

Taking Jet speed (TAS) at the starting point to be 367 knots:

Jet.Frame.Velocity.x = 0
Jet.Frame.Velocity.z = 367 knots

Transform UFO velocity into Jet.Frame (local and world frames are identical at the starting point, so I'll just use one called "Frame"):

Jet.Frame.UFOVelocity (UFO velocity relative to the jet)
Jet.Frame.UFOVelocity.x =-33.28 - 0 = -33.28 knots
Jet.Frame.UFOVelocity.z = 39.52 - 367 = -327 knots

VC = DotProduct(Jet.Frame.UFOVelocity, JetToUFODirection)
VC = -227.2208

Sign is flipped on the FLIR (my negative is their positive, it's just a sign convention thing of no concern here). So looking for a VC of 210 and I get 227. If I didn't make a mistake, I'd call that close enough, especially considering I'm omitting elevation which would reduce the 227 a fair bit and am cloning a vector instead of using the correct value. I was expecting VC to be very far off regardless, but it's not. I only checked the first point, but that one's close enough that I wouldn't be surprised if the other is fine too.

I'm genuinely surprised by this result. Just shows you really do have to sit down and work it out before drawing too many conclusions. I'm also surprised at how well it worked with the simple bank angle/turn rate formula.

Having said that, I am concerned that it's now evident that there's more than one solution that gives the same VC. The FLIR video has a range 10x greater and is also a valid solution with a different speed. Looks to me like they indeed have a solution that works, the trouble is there may be infinitely other solutions that work just as well with different speeds and distances. That could still be useful if it can put a cap on the max and/or min speeds though. Maybe it can after all?

Last edited:
Impeccable.
I ask you: it is possible to take advantage of the parallax of the background to obtain the speed of the object, thus transferring all motions into an absolute frame of reference such as the ground?

Not sure, sorry.

These records are made under completely different conditions. A pelican or bird of prey can reach 3000 meters by taking advantage of thermals, not flying straight uphill for a long time. So also the Indian geese, reach those altitudes but always fly at very low heights from the ground. If a bird was visible in the GOFAST video, there is no biological justification for it to travel at 5000 meters level, over 300 miles from the coast at a constant speed of tens of mph, in winter.....

You're implying that the bar-headed geese over the Himalayas - using thermals to climb that high - can only do so because the mountains continually rise underneath them and thus the birds are never that high off the actual ground? I think you're underestimating the power of thermals... or atmospheric lift, really: (thermals, ridge lift, lee waves, convergence zones).

The unpowered glider altitude record is 76,000 feet, using thermals to achieve that altitude. (Yes, extraordinary conditions.)

Also think about thunderheads over the ocean. Those are water vapor rising in thermals. What altitude do thunderheads achieve?

The high flying bar-headed geese are just one example of high flying birds. I suspect that the ceiling for birds comes when they run out of oxygen, not when they run out of lift.

Winter in Florida. Two hundred miles off the coast of Florida... that's where the Gulf Stream is. Warm water that keeps Florida warm... and spawns thermals? The pelicans achieve 10,000 feet using atmospheric lift and then glide down without flapping to travel to distant feeding areas.

I should start a thread about the history of radar and UFO's. But this is apropos because of the Golden Eagle story at the end of this excerpt.

https://www.nationalarchives.gov.uk/documents/the-ufo-files-extract.pdf

Time frame here is late 40's to late 50's in Great Britain.

One of the features of the UFO phenomenon that most concerned the Air Ministry was visual sightings that appeared to be corroborated by radar operators, as featured in the report by Michael Swiney and David Crofts. Unexplained phenomena had been tracked on RAF radars early in the Second World War and again during the ‘ghost plane’ flap of 1947, but until 1952 none of these had involved visual sightings.
In his history of UK air defense radar systems, Watching the Skies, Jack Gough says that ‘angel’ and ‘ghost’ echoes continued to plague RAF radars during the early 1950s. They sometimes appeared from the ground ‘as a cloud of responses very similar to the echoes obtained by small aircraft’. When tracked as individual echoes they could easily be mistaken for military aircraft as they followed a steady course and were plotted at heights from 2,000 ft to 10,000 ft.
The Air Ministry turned to their scientists to provide a solution to this problem. Initially there were two competing theories to explain ‘angels’. The first was they were caused by unusual conditions in the atmosphere that created pockets of air that bent and reflected radar beams to produce false targets on radars. This appeared likely, but could not explain how some ‘angels’ moved against the prevailing winds or faster than measured wind speeds. The second more improbable-seeming theory was that angels were really formations of birds flying to and from their breeding grounds as part of their annual migrations. At the time the few ornithologists who were using radar to study bird movements had problems persuading the RAF to take this theory seriously. However, during the war staff at coastal radar stations had linked ‘angels’ on their screens with flights of seabirds spotted with the naked eye. On rare occasions large individual birds had been known to cause chaos. Barry Huddart, who served with Fighter Command HQ in 1957, recalled one incident ‘when fighters were scrambled to intercept an echo on a radar screen which turned out to be a Golden Eagle at 25,000 ft in a jet stream...

Last edited:
...there is no biological justification for it to travel at 5000 meters level, over 300 miles from the coast at a constant speed of tens of mph, in winter.....
I do not think the target is likely to have been a bird. That said, while I think it is fair to say that it is unlikely that a bird would be at that altitude at that time and place, it remains possible. Birds, like many other animals, seem to engage in play --in activities "for the Hell of it" just because they might be interesting or fun. Not everything animals do is purely based on strict biological utility.

Some forms of play, called “locomotor play,” seem quite similar to the exhilarating play of children sledding down a steep hill. Some ducks have been observed floating through tidal rapids or fast-moving sections of rivers, and when they’ve reached the end, hurrying back to the beginning to ride over and over. Common Ravens have been observed taking turns sliding down a snowbank on their tails or rolling over and over down a hill. In the air, ravens and crows often rise on air currents only to swoop down toward earth, then glide back upwards, again and again.
Not exactly the same as "how high can I get today?" I admit. But yeah, birds, like the rest of us, sometimes do stuff just for the fun of doing it.

Birds use atmospheric lift to achieve those great altitudes to gain potential energy, and then travel long distances without having to expend energy flapping.

I'm not saying it was a bird, I'm just saying that people don't seem to know much about birds, including the Quantum Physicist Boys.

If I had to bet, I'd bet on an ordinary civilian fixed wing aircraft.

Last edited:
I'm not saying it was a bird, I'm just saying that people don't seem to know much about birds, including the Quantum Physicist Boys.
I'm very familiar with the abilities of birds. But we cannot associate these abilities with all bird species in every place in the world. The Indian geese in their migration fly over the Himalayas at 9000 meters above sea level. But they have not been observed at that altitude on the Deccan Plains. The technique of large birds of prey and vultures is known to exploit thermals to reach high altitudes. This methodology has also been observed in Australian pelicans to first identify large bodies of water. They are rarely observed at 3000 meters. But such extreme behaviors have not been recorded in any Atlantic seabird. There is a vast literature related to the maximum heights reached by seabirds on the Atlantic coasts. The possibility that it is therefore a bird I consider quite remote. Not impossible of course. But it can't be one of the two most obvious solutions, the other a weather balloon, to explain that video.

I'm skeptical of the bird hypothesis. Jets were vectored out to these things by ship radar as far I can tell. While weather radars can see massive flocks of birds (even insects), seeing an individual one (much less tracking a single one) is probably asking a lot. The radar systems will filter out most of the stuff like that anyway. Birds are nothing new. It seems unlikely a carrier strike group would be vectoring F18s out to investigate individual birds and for the pilots to be baffled by what they're seeing upon arrival. If that's really the case, we have a lot bigger problems than UFOs.

Last edited:
Jets were vectored out to these things by ship radar as far I can tell.
Have you got a source for this, I don't think either the pilot or WSO from GoFast ever came forward to give the clips any context?

Even fairly common commercial RADARs have no struggle capturing birds as can be seen from a variety of sources.
From anglers finding fish via. feeding birds (https://www.sportfishingmag.com/boats/marine-electronics/using-radar-find-birds/):
“If I had to pick a single piece of electronics on my boat to go catch fish, without a doubt, it would be my radar,” says Capt. Ed Dwyer, who runs Ticket Sportfishing Charters (321-288-1940, www.ticketfishing.com) out of Port Canaveral
[...]
“I can mark a single frigate bird at six miles,” Dwyer says.
[...]
“We got birds one day at 3.8 miles,” Nugent says. “I can consistently get birds at two-and-a-half to three miles. It’s not an exact science, but 95 percent of the time it works for us. It saves me so much time and so much fuel.”
And from people tracking migrating flocks (https://www.audubon.org/news/how-use-radar-track-birds):
Once you start thinking about it, La Pluma explained, it’s no surprise that radar works for tracking migrating flocks. “These things pick up droplets of water in the atmosphere,” and compared to that, “birds are gigantic targets.”

Since the F18 is using a cone Doppler RADAR, the true size of the target doesn't matter, only it's angular size. If it is able to find an F18 at 27 nmi (which it is more than capable of) then it should be able to find a bird at only 4 nmi. I realise I'm not explaining it that well so here is a sketch to show what I mean (I know we should really be using the RCS instead of wingspans but you get the idea):

Looking at go fast again now I don't think there's a radar lock. If there was he could have slaved the FLIR to it instantly like was done in the gimbal video. In go-fast it looks like the FLIR itself acquires the target (probably looking for moving patterns in contrast or similar) after some manual FLIR steering to try to get it somewhere on screen.

The F18 ATFLIR is often coupled to the AZ/EL RADAR in a way that lets the operator point both the RADAR and the ATFLIR at things while choosing which one to look at, achieving a lock/track on one automatically locks on the other. It might be possible that no one initially found the object on the RADAR but that the operator found it using the ATFLIR and that once they locked it, the coupled RADAR automatically started putting energy down the same line to get the range and closure speed. This way something that the RADAR algorithm would have filtered out normally is now being shown as the RADAR has been specifically pointed at it.

The manual on the coupling between AZ/EL RADAR & FLIR:

I realise a couple of pilots have now said that it could be coming from the ATFLIR 'look angles' but I don't know how this would be possible, surely you would need either the size, speed, altitude or range of the object to calculate that? If what we see in the video is the systems 'best guess' then why has the system made the assumption that this is something small and slow?

On another note, it's good to see a fellow DCS player here. What do you run? I've got the Warthog setup with Oculus. It's mind bending.
I've got the Winwing Orion with Track IR (thankfully for my wallet I prefer seeing the keyboard while playing, can't say I'm not tempted by VR though ).
in DCS I have a hell of a time getting the F18 to hold a radar lock on anything. It's annoying as hell. You can breath too loudly and it'll lose lock. I had a Mig-29 heading straight at me, two miles away and couldn't get the stupid radar to even ping it, much less lock. I have a hard time believing the real plane is that horrible, but if it is, well.... Bird locks at 58/36 degrees 3-4 miles out are probably out of the question.
After many hours of tutorials/reading and practice I've gotten better, still terrible with it though haha. My impression from watching others use it on youtube is that it is a rock solid tool with an incredibly steep learning curve.

Interesting about the bird tracking, thanks for sharing. While that establishes that a bird can be tracked by a hobbyist sport radar system at 6 miles, it doesn't establish anything about the ship's radar interaction with birds because 1) they're different systems and 2) they're designed for different purposes. The interview with a radar operator on one of the ships (think it was not go-fast but same system or similar) said it filters out birds and similar clutter (weather effects too most of the time, he describes what those look like as well), so to me it seems most likely they aren't seeing bird targets. Is it possible they are anyway? Maybe. Is it also possible that a modern radar intended for tracking militarily significant contacts filters non-militarily significant contacts like birds and so forth out? Yes, and we have testimony to that effect. So while I can't claim to know for sure either way, we have speculation on one case with a different device with a different purpose and positive testimony on the relevant system in the other, so I'd have to put more weight on the "ship radar doesn't show birds to the radar operator" possibility whether it can detect and/or track birds or not. Might they get a ping occasionally? Maybe. Are they likely to send jet fighters to intercept a bird? Well..... Like I said before, if our carrier groups are sending jets to intercept birds and the pilots are baffled by what they're seeing upon arrival (especially when there's visual contact off FLIR), we have much more serious problems than UFOs.

Jet IR/radar slave quote from the manual: That says the IR can be slaved to a radar and point the IR system toward a locked radar target, but it doesn't say the reverse is true: That the radar can be swiveled to lock an IR target. Perhaps it says it can somewhere else in the manual? Even if it's not in the DCS manual, it's possible that the real F18 can do it even if the DCS F18 can't (DCS isn't a perfect recreation after all, and if it's classified it better not be in DCS to begin with). If it could I'd think the manual would mention it.

What seems pretty clear is the radar didn't see the thing when it was just a couple of miles away. If it did, they could have locked the radar to it and slaved the FLIR and saved themselves the trouble of hunting around for it manually. If the ship didn't have a radar contact on this one like the other two did, how did the pilots run across this in the first place? By chance? If it's a bird and the F18 radar can see birds, why can't the radar see it? Sounds like we're trying to have it both ways at the same time. Chances are they ran across this not by chance, but the same way the pilots for the other two videos did: Ship radar tracked them and vectored them to intercept. Do I know that for sure in the go fast case? Nope.

Either way we still don't know for sure if the range info is accurate on go fast. I'm leaning toward "no," which is what two out two pilots have said about this. The symbology for laser tracking is missing so we can be pretty sure there's no laser/IR beam on the target, and both pilots say the laser wouldn't be used in this situation anyway. If it's true that the range info is wrong, then there's likely no radar lock either and perhaps the IR/Radar slave mode only works in one direction (radar can put IR sensor on target, but IR sensor can not put radar on target. Manual mentions one but not the other AFAIK unless you've seen otherwise).

On the FLIR look-at angle ranging: I agree with you, geometrically it doesn't make sense that they could determine range from that. I certainly can't from angles alone unless you have two separate aircraft reasonably far apart tracking the object simultaneously so you can triangulate the position. Pilots usually get explanations just deep enough to operate the systems effectively, but that's about it. They don't know the systems anywhere near as well as the engineers do. If I was hired to write targeting pod software to do this, the first thing I'd try is image processing to determine angular size while trying to match the image against a 3D model at all different rotations with AI. If identification probability was too low, I'd call it "unidentified" and try to provide range information by assuming a standard fighter plane size and display that to the pilot. I imagine having some sort of rough range estimate is preferable to nothing.

I think in the FLIR go fast case, that's probably what we're seeing: An estimate derived something along those lines. It's very hard to be accurate with that though, and as you know already, if the actual size isn't the same as the default size, the range info could be off by a long way. If it's half the size the software is guessing it is, the range would be off by 2 miles one way, and if it's double, it'd be off by 4 miles the other, which throws all the analyses (including mine) out the window for the most part. While they're interesting to do and are often informative when they put upper and lower bounds on unknown data (wind speed, jet turn rate, target speed, etc.), we have to take them for what they are:

Each one is one possible solution in a set of infinite possible solutions, nothing more.

This is worth expounding on in detail and ties into confirmation bias on all sides of this issue:

What's bothering me most about all this is people on all sides are simply latching onto whichever solution fits their own wishful thinking and declaring something along the lines of "since analysis X's solution fits explanation X, explanation X is probably right because it is one possible solution to the problem." While it's true that it's a possible solution to the problem and their conclusion could very well be correct, the conclusion is arrived at via fallacious logic because it ignores the rest of the solution space.

For example, if I do an analysis that puts a ground speed of 50 knots on go-fast, then add 50 knots head wind to target to bring it to rest relative to the wind, I now have one solution in which TAS is 0. Therefore "balloon" is a possible solution. While it's true that "balloon" is in the solution space and might be the correct answer, I can not simply stop at this point and say "therefore it is a balloon." I can not even say "therefore it is probably a balloon."

"Probably" equates to "probability." To say "X is probably true" is to claim that there is a higher probability of explanation X being correct than any other explanation. The reason the explanation is not more probable in this example is this: If an unknown 50 knot wind can be inserted in one direction, it can also be inserted in the opposite direction with equal probability because we lack wind information. Now we have a new solution of 100 knots TAS instead of 0. The 100 knot solution is equally probable. We are left with no way to differentiate between "balloon" and "something flying 100 knots TAS into the wind, definitely not a balloon." We need wind data to narrow the solution space enough to draw a definitive conclusion one way or another.

If we are free to simply insert any wind speed we wish, we can go the other direction like I did in my analysis. Jet turn rate could be 0. While I doubt it really is 0, it is possible so therefore the analysis lies within the solution space. So in that analysis I can get 124 knots, and within that I can insert wind speed possibilities as well to generate even more solutions. I don't think anyone would disagree that a 124 knot wind speed difference between those altitudes is out of the realm of possibility, so I can set the wind direction to bring the target air speed to 0 and get a ballon. However, the probability of that wind blowing in the opposite direction is precisely the same, so I could with equal probability say the TAS of the target is 248 knots.

Because the probability of both answers is identical given the available measurements we have access to, we can not accurately say any of the following statements:

1) "TAS is 0 or near 0. It's probably a balloon or bird."
2) "TAS is 248 knots. It's definitely not a balloon or bird"

While they both remain possibilities and one of them may very well be correct given our limited information (not a problem for the Navy, more on that later), we can not say that one is "probably more likely" than the others if the free parameter we're manipulating in the model has an equal probability of facing in either direction and of having any value between 0 and 124 knots or whatever the possible range might be. It could be a headwind or a tail wind. Both are equally probable and therefore we can not pick one solution over the other. The problem gets even worse as you add more free parameters, especially if any have high sensitivity (e.g. turn rate).

Nevertheless, we tend to grab the solution we want to be true and say "since this solution is possible, I know what it "probably" is, so that's good enough for me" and we're done. Nobody on any side of this should be doing that. That's how confirmation bias creeps into analyses like these (including mine) when there is too little information available to draw a definitive conclusion one way or another. I.e., arriving at 100% probability of one solution being correct to another. Then we KNOW the answer. Until then, we don't, so it's unidentified/unknown.

We can go even further than this and use the slower speed analyses that land on 30 knots TAS or so for the target. Stick a 120 knot wind speed difference between the two altitudes into that one and you can make the object fly backwards. So that is also a possible solution and the probability of it being correct is the same as it is for every other solution.

Notice the range of possibilities we can get just using wind speed as the only free parameter. With one parameter I can make it hover, go 250 knots, or even fly backwards. Now add a second or third or fourth free parameter to the model. Add unknown range, unknown turn rate, and whatever else is missing to really know for sure what it is, and I can turn go-fast into almost anything I want it to be.

John von Neumann famously said:

With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

It gets us no closer to knowing what is "probably right" because we are missing all of the free parameter information necessary to reduce the solution space enough to positively identify the object. Maybe we can establish that range is accurate (not looking good so far) so we can eliminate one free parameter. Maybe we can find winds aloft data in the area and eliminate another. Maybe we can get turn rate telemetry and eliminate another.

Now here's the kicker and why we should take the "unknown" status of these videos rather seriously: Since this video is now officially in their "unknown" bucket, they have done their own analysis. ATIP head who is charged with this said they in fact did do so. They do not have the same free parameter problems the rest of us do. Navy/Pentagon/US intelligence/whatever has (or had, I don't know how long information like that is kept) all of that information because everything that can be recorded on a flight is recorded. They will have the precise temperature of the left rear compressor at 1:02:25.21 into the flight along with every other telemetry output.

Here's what this means for free parameters and why their analysis is so much more significant than mine or anyone else's:

They will have a precise GPS track of the jet and precise turn rate information, so do not need to guess at the flight path like we do. That eliminates turn rate as a free parameter. That narrows the solution space.

They will know the exact status and accuracy of the FLIR range readout and will have whatever radar data is available from not one, but two aircraft and the ship's radar. That eliminates range as a free parameter and narrows the solution space even further.

Remember the ships vectored out the jets with their radars in at least two of these videos, so there is ship radar information which we don't see. So they'll have precise radar tracks on at least one or two of the objects which can be cross correlated with the information on the jet (FLIR pod and so on), multiple sources for the same variable, so even the ground speed of the targets is probably known to them at every point long before and after the videos. Corbell's ship radar video demonstrates this to be the case and is confirmed by testimony from one of the ships radar operators involved in one of the three videos. They can simply measure target velocity with ship radar directly. That eliminates target speed information as a free parameter and narrows the solution space even further.

What this means is they have a heck of a lot more information than we do and a lot less free parameters to fiddle with in their analyses than the rest of us have. They can simply measure all the values we are stuck guessing. Their solution space is much narrower as a result. The probabilities of each possible solution being the correct one are narrowed down more than we could ever hope for, so they are able to rule out possibilities when the rest of us can not. If they could determine from their telemetry that the TAS is not 0, then they could have definitively ruled out the balloon hypothesis even before any of us saw the video. I would think they would have categorized the video accordingly. They didn't. They categorized it as "unknown."

Now here's what keeps me up at night: Their conclusion is not "bird" or "balloon" or "airplane" despite having none of the free parameters that we have. To them it's "unknown." I don't know if that means they were unable to decide on what it was or if they were actually able to rule those out of the possibilities. That would be very interesting to know. While it's fun to make these models and give it our best shot, none of us are going to do any better than they did. We really just have to say "we don't know the correct solution for what we see in these videos" and leave it at that. To us out here in internet land, a balloon is one of many possibilities, but to them it apparently wasn't, or they weren't able to determine definitively that it was even with all the telemetry info the rest of us are lacking. We should not be too quick to dismiss that fact and be too confident about our own analyses, thinking we've outsmarted US Intelligence by finding solutions via manipulation of free parameters that they simply don't have to deal with at all in theirs.

So what are these things? The only rational conclusion we can draw given the available information in the videos is "unknown."

Last edited:
- Yes, early radar was difficult to use because it picked up all sorts of irrelevant clutter. Steps were taken to electronically tune out as much clutter as possible.

-There's a long, extensive history of people seeing things in the sky they can't identify then asking someone else to look at radar to see if there's anything they can't identify, then falsely concluding they are looking at the same thing(s). E.g. Saucers over Washington D.C. 1952, The Kaikoura Lights 1978, The Rendlesham Forest UFO 1980, etc., etc.

-Currently, (and since the 1949) there are things floating around the sky that are meant to be seen on radar.

-Twenty-first century radar is still far from infallible.

What are the various things RADAR is intended to detect?
RADAR is intended mainly to detect aircraft, including unmanned aircraft and missiles.

RADAR is also used to detect ships, weather, weather balloons with RADAR targets, satellites, and other miscellaneous items.
A special kind of RADAR (called Doppler RADAR) is used to detect the speeds of objects toward or away from the RADAR set. Examples of these objects include motor vehicles, weather objects, bullets, baseballs, and track-and-field equipment. It can also detect the rotation of a tornado.
• Does RADAR detect things it is not supposed to?
Yes. RADAR can detect all of the following, and many more:
1. Metal bird identification tags
2. Flocks of birds
3. Swarms of insects
4. Objects on the ground
5. Tall buildings
6. Mountain peaks
7. Corner reflectors (e.g. a dump truck or railroad gondola hopper car)*
8. Some model aircraft
9. Satellites
10. Meteors and re-entry events
11. Some kinds of prank UFO balloons [with makeshift aluminum foil targets made by knowledgeable pranksters]
12. Objects detected through anomalous propagation
13. Some objects beyond the horizon
14. * Does not have to be in the strongest part of the beam to be detected, so the displayed direction could be wrong.
15. Doppler RADAR has its own list of objects it is not supposed to detect. These objects or effects cause the reported speed to be wrong:
16. Rotating or vibrating parts on a vehicle or aircraft
17. Wind turbine power generators
18. Spinning or flapping decorative devices (such as flapping silvered flags at a used car lot)
19. Flapping tree leaves blowing in the wind
20. Signs vibrating in the wind
21. Windblown rain, sleet, or snow (when not intended to be used for weather)
22. Vehicles not intended to be tracked (e.g. aircraft in the distance)
23. Satellites
24. Meteors and re-entry events
25. Objects detected through anomalous propagation
26. Some objects beyond the horizon.

During the early days of RADAR in World War II and the next few years, the causes of unknown targets on RADAR were not known. Someone joked that they might be tracking angels, and the name stuck.
The unknown targets were anomalous targets, which have various causes.
[*]What kinds of effects cause anomalous targets on RADAR?
There are several different effects that cause unknown targets to appear on RADARscopes when objects are not actually at those locations:
• Multiple-Reflection Beam Path
The RADAR beam bounces off more than one object before returning to the RADAR set. It produces a target that appears farther away than any of the real objects in the multiple reflection.
• Multiple-Trip Echo
The RADAR beam bounces off something that is so far away that the set has sent out more pulses before the echo comes back. It is displayed as though it belongs to a later pulse, much closer than the real object is.
• Corner Reflector
Leakage from the RADAR antenna near the beam strikes an efficient corner reflector (a place where three metal surfaces all meet at right angles, making an inside corner). The corner reflector then reflects this leakage back to the RADAR antenna, producing a signal strong enough to make a blip.
Objects that make good corner reflectors include rectangular dumpsters, dump trucks, open-back box trucks, and railroad hopper cars.
The corner reflector can produce a blip in the wrong compass direction, and it can participate in a multiple reflection (above) or a side lobe blip (below).
• Side Lobe Blip
Some of the RADAR beam leaks out of the antenna in the wrong direction due to the shape of the antenna. A very reflective target can produce a blip when the energy of the side lobe of the antenna hits it. That object makes multiple blips, all the same distance from the set.
• Tall Objects on the Ground
Objects tall enough to be struck by the RADAR beam will appear as blips on the screen. These include tall buildings and antenna towers. But they are easy to identify on the screen, because they never move.
• Ships on the Water
The RADAR beam can get close enough to the water within the closest 50 miles to pick up ships. The interface between a broadside ship and the water also makes a good corner reflector.
• Ground Clutter
Within the first few miles of the RADAR set, many objects on the ground are struck by the beam and produce blips, causing a continuous large target called ground clutter. Objects cannot be tracked through ground clutter.
• Small Object
A small object that is present at the location shown by the RADAR can produce the RADAR blip without being seen visually by ground or air observers. Such objects include metal-tagged birds, escaped aluminized Mylar toy balloons, and prank UFO balloons.
• Weather
In addition to producing areas of weather clutter, a concentrated core in a storm can produce a blip similar to one made by an aircraft.
In wartime, RADAR jammers are transmitters that usually cover the screen with lines, dots, or a continuous smear that makes the RADAR set useless for detecting the enemy.
• Chaff, and Window
In wartime, chaff (also known as window) is metal foil dropped from aircraft or missiles that covers the RADAR screen with blips. The RADAR operator can't tell which blips are real aircraft and which are chaff.
• Interference from Electrical and Electronic Devices
Electrical devices can produce unwanted radiation at RADAR frequencies, especially if something goes wrong. Electrical shorts in power lines give off broadband blasts of energy, causing random blips that do not repeat. Other devices inadvertently producing RADAR frequencies usually do not make sharp blips. They make smeary patches on the screen.
• Interference from another RADAR Set on the Same Frequency
Anomalous propagation can cause interference between two RADAR sets that are normally far enough apart that they don't affect each other. Such blips usually jump around the screen because the sets are not synchronized with each other. Operators have called these targets "running rabbits."
• Anomalous Propagation
Anomalous propagation is an atmospheric condition that bends RADAR beams. This causes objects that would not normally appear on RADAR to show up on the screen. Often the RADAR beam bends downward, striking objects that are on the ground, are on the ocean, or are normally beyond the RADAR horizon.
• Increased Gain Objects
If the RADAR receiver gain (or the AGC gain) is inadvertently turned up too high, many marginal RADAR targets can appear as planelike blips. These include birds, swarms of insects, weather targets, and other objects that normally do not display as blips.
[*]What is anomalous propagation?
Anomalous propagation is an atmospheric condition where the RADAR beams, instead of moving in straight lines, follow curved or bent paths. This causes objects that the beam would not normally strike to show up on the screen. It works similarly to a visual mirage, and is caused by the same atmospheric conditions.
[*]What causes anomalous propagation?
Anomalous propagation occurs when different layers of air with different temperatures and humidities come together to distort the paths of light rays and RADAR beams. Often a temperature inversion causes anomalous propagation. When stars near the horizon appear to dance around and change colors, it is a sign that anomalous propagation is occurring.
Anomalous propagation can either reflect or refract the RADAR beam, causing it to go in an unexpected direction. Usually it is bent down, where it strikes objects on the ground and aircraft beyond the RADAR horizon. Sometimes it focuses the beam, enabling it to pick up objects far beyond the normal range of the RADAR set.
[*]What is a temperature inversion?
A temperature inversion is a layer of reversal of the normal decrease in temperature with an increase in altitude. It can cause light rays and RADAR beams to bend.
[*]What are the various kinds of RADAR?
There are several kinds of RADAR intended for several different purposes:
• Surveillance RADAR is used to keep track of aircraft in a general area and prevent collisions.
• Approach Control RADAR is used to direct a landing aircraft safely to the runway.
• Ground RADAR monitors all aircraft that are on runways and taxiways on the ground at an airport.
• Tactical RADAR is used to keep track of the relative position of hostile aircraft. It usually has a very quick scan rate and a height finder.
• Nautical RADAR is used by ship pilots to see and avoid other ships in the area.
• Astronomical RADAR is used to accurately measure the position of an object in space or map its surface.
• Satellite Tracking RADAR is used to track the orbits of satellites.
• Weather RADAR is used to find high concentrations of water vapor that make up storms.
• Doppler Weather RADAR is used to find the motion of storms.
• Police Doppler RADAR is used to measure the speeds of motor vehicles.
• Sports Doppler RADAR is used to measure the speeds of objects that were thrown or hit.
[*]What is MTI?
MTI stands for Moving Target Indicator. This is a system devised in the 1950s to remove all stationary targets from the RADAR screen. It originally worked using a mercury delay line to delay each returned RADAR reflection the exact time between transmitted pulses. If a returned reflection appears at exactly the same time as the previous returned reflection, the blip so produced is removed.
Newer RADAR sets use computers to accomplish the same result. That result is to remove any target that is not moving from the RADAR screen.
[*]What can make a stationary target persist on an MTI screen?
There are several different effects that can cause this:
• An object that is spinning (e.g. a helicopter rotor) or vibrating can fool the MTI device.
• An object moving in a tight circle may appear to be stationary on the RADAR, but it is not.
• A target caused by anomalous propagation can move to different points on the ground.
• A target caused by anomalous propagation can rapidly appear and disappear, fooling the MTI.
• A weather target can rapidly appear and disappear or shift its location around, fooling the MTI.
Primary RADAR is RADAR that displays every echo returned to it as a blip. It detects aircraft by detecting a pulse from the RADAR set that is reflected from the metal parts of the aircraft. The aircraft needs no special equipment to be detected.

### Are Some Of The UFOs Navy Pilots Are Encountering Actually Airborne Radar Reflectors?​

The description given of objects involved in numerous close UFO encounters with Navy pilots off America's eastern seaboard during the 2014-2015 timeframe is akin to a 'beachball' or orb with a cube suspended inside of it with the cube's corners touching or nearly touching its edges. This sounds amazingly bizarre and is more reminiscent of what we would expect from a sci-fi movie circa the 1980s than the classic flying saucer or even the large 'Tic Tac' that Navy pilots encountered back in 2004, but to me, it also sounded eerily familiar.
When I thought of round orbs with cubes inside them, balloons and radar reflector devices came immediately to mind. I began hashing out this possibility with my colleague Joseph Trevithick shortly after the reports came to light. The reality is that traditional high-altitude balloons and radar reflectors already go hand-in-hand.

Because a high-altitude balloon doesn't have much, if any, of a radar cross-section, metallic radar reflectors, which come in a variety of geometric shapes, are strung below its gas envelope, thus providing a radar return so that it can be tracked. The combination can look pretty bizarre in and of itself and they are cumbersome and clumsy arrangements. But couldn't this be simplified for more conducive deployment and better aerodynamics by just suspending the reflector inside the balloon itself? A similar arrangement is used for radar reflectors that float on the water or are strung up on ships, but what about one that has to travel through the atmosphere?
Just as I thought, an answer to that question has already been proposed. After searching sporadically over a number of days for what I envisioned in my mind, I found just that in U.S. Patent #2,463,517, titled "Airborne Corner Reflector."

USPTO
Airborne Corner Reflector.
The patent was filed way back in 1945 and was granted in 1949. It is alarmingly similar in appearance to what the pilots had reported seeing multiple times over the Atlantic Ocean. In fact, a near miss encounter with one of these objects as described by Navy Super Hornet pilot Ryan Graves states that the object was likely standing still, floating in the air, when the Super Hornet blasted by at a too close for comfort distance. In other words, it wasn't making any extreme performance maneuvers while within visual range. Instead, it was acting like, well, a balloon.
Hear Ryan Graves describe the encounter in a clip from History Channel's To The Stars Academy-helmed show Unidentified:

Source: https://youtu.be/Oz8k8Iqx3wA

Other statements from Graves and a squadronmate have pointed to the fact that these objects can stay in the air for many hours at a time. This is a characteristic also possessed by a balloon of some sort. Even the perceived threat from a collision with one of the objects and the Navy's lack of interest in dealing with it at the time wouldn't be as surprising as it is now if they were indeed balloons. Weather balloons and other high-altitude balloons are launched into the skies daily and fly among airliners without the ability to track or avoid them. You can read more about this reality here.

Last edited:
Neat information, thanks. Couple questions:

1) What was the airspeed of go-fast?

2) Was it locked on the plane's radar?

Neat information, thanks. Couple questions:

1) What was the airspeed of go-fast?

2) Was it locked on the plane's radar?

I feel you are repeating yourself, didn't we just have this discussion?

I feel I am not. We haven't finished the discussion. I felt the reply to my last post was essentially "radar reflecting balloons exist, therefore go-fast was a radar balloon." I felt as though the entirety of everything I'd just said had been dismissed, ignored. That isn't discussion.

My reply was meant to prompt thought into the possibility that without that information it is not logically possible to come to that conclusion given the information available, even if the conclusion turned out to be correct in the end. If that is unacceptable behavior or my posts annoy you, I can find my way out.

Last edited:
I feel I am not. We haven't finished the discussion. I felt the reply to my last post was essentially "radar reflecting balloons exist, therefore go-fast was a radar balloon." I felt as though the entirety of everything I'd just said had been dismissed, ignored. That isn't discussion.

My reply was meant to prompt thought into the possibility that without that information it is not logically possible to come to that conclusion given the information available, even if the conclusion turned out to be correct in the end. If that is unacceptable behavior or my posts annoy you, I can find my way out.
Thank you very much for your effort. In my first post published in this blog I stated the same thing. It isn't possible to identify the object because the most important data of all, namely the upper air data, is missing. Being aware of the wind intensity at UAP altitude makes all the difference in the world between a self-propelled object and one that is not. As an aviation security analyst, I couldn't help but ask myself an infinite number of questions, including the one according to which the background you see in GOFAST could be a cloud layer. For me it was a shock that Mick West naively declared in his video: "Explained! ..." Many users feel compelled to continue this crusade by proposing, in my opinion, euphemistically original hypotheses. A radar detector balloon dated 1945, or an Australian pelican on vacation in the seas of the East Coast. I think the survey carried out by the user Dimebag2 is very interesting. Most users are convinced that GOFAST has been debunked. There is a fear that if one doubts, he automatically becomes a follower of George Adamsky or Bill Meier.
The Pentagon report ruled that out of 144 cases only one has been resolved. It means that at least 2 of these videos are of unknown origin. What fascinates me is precisely to discover some characteristics by reasoning and applying mathematics as you and other talented users of this site did.

The "bunk" is Go Fast is a very fast and low flying object. The implication of this bunk is that this object is some sort of unknown technology.

The debunk is that the figures shown on the screen show that it is not in fact a very fast and very low object, and that it is operating within parameters that do not exclude a raft of known objects.

As to what it actually is we don't know, but have proposed some possibilities after having debunked the original claims.

Last edited:
A radar detector balloon dated 1945, or an Australian pelican on vacation in the seas of the East Coast.
Wut? This is very muddled.

I agree with the fact that we are missing some critical data to give any strong conclusion here (the wind shear especially). But if using different methods, we find convergence in the results, that would suggest the results are not so off.

I went back reading earlier posts from this thread, and around pages 5-6 there are very interesting analyses that use the first segment of the video, when the object is not locked, to derive the speed. It is based on the 1.8 sec when the object crosses the field of view (FOV) at mark 0'05 of the video. Knowing the FOV, we can estimate the distance covered, hence the speed of GoFast. This is a crude estimate because based on 1.8sec of footage only, but it has the advantage of removing one source of uncertainty, the plane trajectory. In this method, the apparent speed due to parallax has to be accounted for and removed from the GoFast speed. The formula for the parallax speed is Vpar=Vjet*(HGofast/Hjet) with Vjet the speed of the plane, HGofast and Hjet the altitude of GoFast and of the plane respectively (see this post, https://www.metabunk.org/threads/go...e-stars-academy-bird-balloon.9569/post-220457)

I have tried to put this in a 3D geometry model, so we can play with the altitude of GoFast for a given FOV, to see what speed we obtain at different altitudes. What I do is I construct a cone that represents the field of view. FOV can be adjusted in the parameters, I use 0.7° as suggested previously in the thread. Drawing a plan parallel to the ocean that intersect with the cone, I have two points that represent the distance covered by GoFast in the field of view, for a trajectory loosely parallel to the plane heading. I can then calculate the speed of Gofast based on this covered distance, after removing the apparent speed due to parallax (0 if GoFast is at the surface, equals to Vjet next to the jet).

https://www.geogebra.org/3d/cfsvvzxw

My idea is to check if the speeds I get compare with the results from my first model, that is based on a completely different approach when the object is locked (using plane trajectory and camera angles): https://www.geogebra.org/m/axynssmq

I don't know if this is pure luck, but I find it interesting that I get quite similar results for GoFast being at an altitude of 2.1 Nm (i.e. altitude that it would have if the indicated slant range is correct). I retrieve a speed of ~60 Knots, like in my other model. Of course this is based on a FOV of 0.7°, the speeds are much higher for a larger FOV, so I hope 0.7 is correct. Near the surface there are larger discrepancies between the two models.

I appreciate any feedback and correct me if I'm wrong in that analysis, I just made this and may be missing important things. The key parameter here is the FOV, so please let me know if you think 0.7° is a good guess.

EDIT : like always with this kind of analyses, small changes in parameters (FOV, altitude) make significant differences. So take them with a grain of salt, it shows that at least it's in the ballpark of previous estimates. At least this model illustrates the impact of FOV and parallax in function of altitude.

Last edited:
Wut? This is very muddled.

I don't think I'm confused. This is a pre-Cold War object. Unless he's still hovering in Virginia skies for 75 years ....
I repeat.
If you testify to me through a documented observation that there are Atlantic seabirds that usually soar at 5000 meters, I am happy to consider the hypothesis of a bird as very probable.

Last edited:

I don't think I'm confused. This is a pre-Cold War object. Unless he's still hovering in Virginia skies for 75 years ....
I repeat.
If you testify to me through a documented observation that there are Atlantic seabirds that usually soar at 5000 meters, I am happy to consider the hypothesis of a bird as very probable.
The FAA mandates radar reflectors on high altitude balloons in most places in the US.

The FAA mandates radar reflectors on high altitude balloons in most places in the US.
Surely. But the problem is that an association has been made between what some fighter pilots saw and an object that is no longer in use. What the FAA sends is nothing like what they saw.

Surely. But the problem is that an association has been made between what some fighter pilots saw and an object that is no longer in use. What the FAA sends is nothing like what they saw.

What are you talking about? No one saw the Go Fast object other than as small cold blob on the ATFLIR.

I agree with the fact that we are missing some critical data to give any strong conclusion here (the wind shear especially). But if using different methods, we find convergence in the results, that would suggest the results are not so off.

I went back reading earlier posts from this thread, and around pages 5-6 there are very interesting analyses that use the first segment of the video, when the object is not locked, to derive the speed. It is based on the 1.8 sec when the object crosses the field of view (FOV) at mark 0'05 of the video. Knowing the FOV, we can estimate the distance covered, hence the speed of GoFast. This is a crude estimate because based on 1.8sec of footage only, but it has the advantage of removing one source of uncertainty, the plane trajectory. In this method, the apparent speed due to parallax has to be accounted for and removed from the GoFast speed. The formula for the parallax speed is Vpar=Vjet*(HGofast/Hjet) with Vjet the speed of the plane, HGofast and Hjet the altitude of GoFast and of the plane respectively (see this post, https://www.metabunk.org/threads/go...e-stars-academy-bird-balloon.9569/post-220457)

I have tried to put this in a 3D geometry model, so we can play with the altitude of GoFast for a given FOV, to see what speed we obtain at different altitudes. What I do is I construct a cone that represents the field of view. FOV can be adjusted in the parameters, I use 0.7° as suggested previously in the thread. Drawing a plan parallel to the ocean that intersect with the cone, I have two points that represent the distance covered by GoFast in the field of view, for a trajectory loosely parallel to the plane heading. I can then calculate the speed of Gofast based on this covered distance, after removing the apparent speed due to parallax (0 if GoFast is at the surface, equals to Vjet next to the jet).

https://www.geogebra.org/3d/cfsvvzxw

My idea is to check if the speeds I get compare with the results from my first model, that is based on a completely different approach when the object is locked (using plane trajectory and camera angles): https://www.geogebra.org/m/axynssmq

I don't know if this is pure luck, but I find it interesting that I get quite similar results for GoFast being at an altitude of 2.1 Nm (i.e. altitude that it would have if the indicated slant range is correct). I retrieve a speed of ~60 Knots, like in my other model. Of course this is based on a FOV of 0.7°, the speeds are much higher for a larger FOV, so I hope 0.7 is correct. Near the surface there are larger discrepancies between the two models.

I appreciate any feedback and correct me if I'm wrong in that analysis, I just made this and may be missing important things. The key parameter here is the FOV, so please let me know if you think 0.7° is a good guess.

EDIT : like always with this kind of analyses, small changes in parameters (FOV, altitude) make significant differences. So take them with a grain of salt, it shows that at least it's in the ballpark of previous estimates. At least this model illustrates the impact of FOV and parallax in function of altitude.
Really interesting!
A few questions before analyzing your study.
I have read in some manuals that the Targeting Pod without designation remains with a constant AZ / EL during snowplow pointing mode. Have you applied this method?
If that were the case, it would be easy to determine the ground speed of the Fighter and the object with a simple measurement of the angular velocity of the parallax. In your opinion is it possible, or does some parameter take over that I have forgotten?

I am referring to the conclusions of this article posted by Z.F.Wolf and referring in general to the swarm of observations of 2014-2015 and not in particular to GOFAST.

Generally, I think you'll find most people on here tend to focus on single events and single claims about those events to debunk. So with GOFAST the purpose isn't to identify the object, it is to debunk (or verify) the claim that the video shows something near the water moving at a high rate of speed. Speculating on events over time as a singular hypothetical phenomenon isn't really in line with the purpose of the forum IMO.

Generally, I think you'll find most people on here tend to focus on single events and single claims about those events to debunk. So with GOFAST the purpose isn't to identify the object, it is to debunk (or verify) the claim that the video shows something near the water moving at a high rate of speed. Speculating on events over time as a singular hypothetical phenomenon isn't really in line with the purpose of the forum IMO.
Unfortunately I don't know well the story of the publication of these footages. But I wonder if really the US Navy declared a note attached to the release in which it spoke of a very low and fast object or is it a statement derived from the many that have said the various Elizondo, Mellon, Corbell, TTSA etc.etc. Because if the latter were true, here we are debunking a personal affirmation. So conceptually it cannot be brought to a certain level.

Really interesting!
A few questions before analyzing your study.
I have read in some manuals that the Targeting Pod without designation remains with a constant AZ / EL during snowplow pointing mode. Have you applied this method?
If that were the case, it would be easy to determine the ground speed of the Fighter and the object with a simple measurement of the angular velocity of the parallax. In your opinion is it possible, or does some parameter take over that I have forgotten?
Thanks Leonardo. Yes this is exactly what I'm trying to do here. Because without locking (snowplow pointing mode?) the camera angles do not change (or not much), the speed of GoFast can be estimated from the time it takes to cross the field of view. And for a given altitude of GoFast, knowing the plane speed and the line of bearing with GoFast (-24°), we can estimate the apparent speed due to parallax alone, and remove it from the previous number. My model gives all the potential speeds along the altitude, based on this calculation.

If I understand correctly what has been posted before this is how the TTSA estimated the object speed originally (?). Depending on the FOV you use the speed can get pretty insane near the surface (230 knots for a FOV of 0.7, more than 500 knots for a FOV of 1.5).

Many users feel compelled to continue this crusade by proposing, in my opinion, euphemistically original hypotheses.... There is a fear that if one doubts, he automatically becomes a follower of George Adamsky or Bill Meier.

I'd go easy on assuming thoughts and motives to other users. I'd especailly be cautious about that when it might come off as a straw man as opposed to just debating the interpretations of the data.

Unfortunately I don't know well the story of the publication of these footages. But I wonder if really the US Navy declared a note attached to the release in which it spoke of a very low and fast object or is it a statement derived from the many that have said the various Elizondo, Mellon, TTSA etc.etc. Because if the latter were true, here we are debunking a personal affirmation. So conceptually it cannot be brought to a certain level.

Thanks Leonardo. Yes this is exactly what I'm trying to do here. Because without locking (snowplow pointing mode?) the camera angles do not change (or not much), the speed of GoFast can be estimated from the time it takes to cross the field of view. And for a given altitude of GoFast, knowing the plane speed and the line of bearing with GoFast (-24°), we can estimate the apparent speed due to parallax alone, and remove it from the previous number. My model gives all the potential speeds along the altitude, based on this calculation.

If I understand correctly what has been posted before this is how the TTSA estimated the object speed originally (?). Depending on the FOV you use the speed can get pretty insane near the surface (230 knots for a FOV of 0.7, more than 500 knots for a FOV of 1.5).
Perfect! I will try to do similar work with 3d software and then compare the results. I have seen that there is also another piece of footage at about sixth seconds that can be analyzed to compare the data. What do you say?

@Leonardo Cuellar , yes that's the one I'm using, between 0'05 and 0'07. A user had looked at each frame and counted 1.8sec for the object to cross the FOV, based on frame rate.

@Leonardo Cuellar , yes that's the one I'm using, between 0'05 and 0'07. A user had looked at each frame and counted 1.8sec for the object to cross the FOV, based on frame rate.
Excuse me. I thought you were referring to the previous fragment around 0'02.
I had also initiated an investigation into the steering cue dot to understand its behavior. I finally understood how it works. From the calculations made it would seem that this can also provide us with the drift angle and the actual direction of the wind as it is dependent on a frame of reference where the axes are those of the fighter and not of the environment.
Taking a quick look at the data it would seem to me that the fighter had all the headwind and then became slightly crosswind during the turn. This may later be useful in determining a reliable airspeed of the UAP.

In DCS I've always understood the dot to be a bearing indicator that's there for a quick "it's at 10 o'clock" type of check. I.e., it's reproducing the bearing info as a quick visual reference. Top of FLIR means 12 o'clock, right means 3 o'clock, bottom is 6 o'clock, etc.. If so I wouldn't think wind speed could be determined from that. Maybe the dot shows more information than I thought I knew. I understood the distance from the origin to represent elevation: Center means "down" or "up" and "far away from center" means horizontal plane. In that case it's simply reproducing the numerical elevation/bearing we already have. Am I mistaken?

"The original debunk" may have been to show it not being an extraordinarily fast object. That's fine, but are we restricted to only discussing a top speed limit here or can we also discuss balloons/birds and so forth?

I see no way of determining either ground speed or TAS of the target with any accuracy given the info we have which was the point of my huge post earlier. The best we can do is very roughly estimate the speed in the jet's frame of reference, and even that is very dodgy because we don't know range or the jet's turn rate. As Mick and others have shown, the turn rate is an extremely sensitive parameter in all this. As a vehicle dynamics guy, I don't accept that bank angle gives you an accurate turn rate even in a constant altitude state. The simple formula everyone's using there is a 2D solution that only holds true under a specific set of constraints, including:

1) Lift vector is the only force involved aside from gravity, or all other forces cancel, leaving lift vector as the only parameter. This is not true in a real aircraft.

2) The lift vector is aligned in the vertical plane cutting through the turn center (turn center isn't constant anyway). This is not true in the presence of any angle of attack because the lift vector rotates backwards away from this plane. What Lehto said about aircraft loading in his first video which Mick subsequently ignored was correct. The load matters because it affects the AOA which affects the rotation of this lift vector. If the lift vector requirements for level flight at any given bank angle change due to this effect, so will the observed turn rate. This rotation is the primary reason for the substantial induced drag on jets when they turn, and is why you need ridiculous thrust levels to maintain speed in high g turns. It is not insignificant. The engine thrust is literally designed around this fact. The F22 flight computer's crazy control surface operation where they even get the ailerons and spoilers/flaps involved in steady state turns is there in order to minimize this lift vector rotation in turns. The control surfaces operation is not simply changing yaw/pitch/roll moments as it is in a small private plane without a fly-by-wire system. The control surfaces are deliberately manipulating the location of center of lift with respect to the center of gravity (calculated in real time by the computer according to loadout, fuel levels, etc) in order to maximize plan view area and use the lift vector location for pitch rate moment control so control surface deflection can be minimized (further reducing drag), all with the aim of decreasing this lift vector rotation (AOA) as much as possible, thereby reducing induced drag and maximizing turn rate at any given angle of attack. They would not bother doing all this if the simple bank angle formula was all that determined turn rate. Real flight is a hell of a lot more complicated than most of the people doing these simple analyses realize. They spend a minute googling a formula, think it's some law of physics and the only parameter involved, and run with it. Next thing you know they think they've proven the target was a radar reflected balloon or bird or something. They haven't. They don't know what they don't know. They are running too far outside their area of expertise . As scientists, they should know better. This also true in any cases where people try to estimate turn rate from the g-meter on the HUD in some of the other videos. What they don't realize is that is a different frame of reference from the actual turn. If you roll the plane 90 degrees and pitch it 90, you'll get huge g and 0 turn rate. They aren't aerospace or simulation engineers and quite frankly don't know what the hell they're talking about. Worse, they don't even realize it.

3) The thrust vector has no vertical or horizontal component relative to flight path direction. This is not true in a real aircraft.

4) There are no lateral forces on the jet in the jet's local frame, which in a banked turn will have both vertical and horizontal global components in the aforementioned turn centered reference frame. This is not true in a real aircraft either, especially one like an F18 with a fly-by-wire system, even more so if any of the autopilot systems are running like the altitude hold which appears to be the case in all three of the videos. There is side slip and control surface modulation being done to manipulate the yaw/pitch/roll moments throughout all this, all of which induce side forces. Side forces are not 0. Period.

All four of these change the requirements on the lift vector to maintain constant altitude flight at any given bank angle, therefore they change the turn rate results at that same bank angle. Unknown pilot or autopilot/flight computer control inputs make this even worse. While the simple bank angle equation is a reasonable starting point and good for some quick, back of the napkin estimates, that is all it is good for. It is in no way definitive. No aerospace engineer will be using that as the be all/end all when predicting turn rate in anything other than simple back of the math approximations prior to a real analysis or simulation run in 3D.

We do not know the jet's turn rate. End of story.

Convergence of multiple analyses on a small range of answers: This is not of any significance. It would be expected if everyone is using roughly the same values for the free parameters, which again is expected if everyone is doing their own analyses based on the work of others'. The point I was trying to make with my initial 0 turn rate analysis was that using a valid value for turn rate it is possible to get a very different result from others. I wasn't trying to argue that my analysis gave the one and only correct answer, only to show that there is a huge range of answers we can get by fiddling with all the free parameters we have to play with here.

The biggest problem in all this isn't even the hyper sensitive turn rate. It's wind velocity. We have two of these, one at the target altitude and another at the jet's. Both are unknown and each could have a range of perhaps 0-130 knots independently in different directions, a potential difference in wind speed of 260 knots between the two altitudes at the most extreme. I doubt it's really that much, but we basically have a couple hundred knots to play with. Whatever analysis anyone comes up with, we are free to change the target TAS by that amount.

It's fine to play around with this, but we shouldn't pay much attention to the fact that many people are converging on roughly the same answer, especially when those doing so often appear more interested in "proving" it's a balloon or bird than actually exploring the entire solution space to see what can and can not be ruled out, and in most cases are just duplicating previous results with slightly different free parameter values. It's no wonder they're similar. My larger point was simply showing the faulty logic in the "if I can show a possible solution that fits explanation X, I know that is the correct explanation." Most people on both sides of this issue seem to be doing that. The Navy or ATIP or whoever analyzed this with the full set of data doesn't have that free parameter problem and have crossed "radar reflected balloon" off of their list as a definite "no" according to ATIP head. Many are ignoring that completely and holding on to the balloon idea anyway for whatever reason, as if to say "radar reflective balloons exist, balloons move slowly, our target *might* be moving slowly, therefore the target is a radar reflected balloon." Nonsense.

Last edited:

I don't think I'm confused. This is a pre-Cold War object. Unless he's still hovering in Virginia skies for 75 years ....
I repeat.
If you testify to me through a documented observation that there are Atlantic seabirds that usually soar at 5000 meters, I am happy to consider the hypothesis of a bird as very probable.
Florida, not Virginia.

Because this is an illustration from 1949 you're implying the author of this article - and I - are so confused that we think something from 1949 is still floating around? The author of the article used a 1949 illustration because photos of airborne corner reflectors currently in use are hard to find on the Internet. It's hard for me to believe you don't understand that.

I have found a photo of ...

https://www.irvingq.com/irvingq-pro...s-corner-reflector/floating-decoy-system-fds/

IrvinGQ’ FDS3 passive off-board corner reflector floating decoy system is a ship-deployed surface floating passive RF soft kill countermeasure. It has been designed to protect naval ships from RF-seeking missile threats...

There's a corner reflector array inside a plastic envelope. Very much like a corner reflector cluster inside a balloon. (No, the GoFast UFO was not something floating on the surface of the ocean offshore of Florida.)

From 1978 (No, they're not still floating around the skies of Florida.)

A hobbyist made this one. (No, this is not the GoFast UFO.)

To belabor the point...

This is what a bus looked like in 1949. They still look pretty much like this.

1949 GMC Model PDA-4101

(No, I'm not saying the GoFast UFO was a 1949 GMC bus floating around in the sky offshore of Virginia Florida.)

Last edited:
I'm satisfied that the one pilot describing a cube in a sphere was probably one of these odd balloons, so I appreciate you sharing that because I wasn't aware of the bizarre shapes these things have. However, I don't think it's relevant to what I was discussing: Go-fast. ATIP ruled out balloon as an explanation for this one. If they have full radar data in their analysis (which they most certainly will), all they need is two pieces of information to rule out balloons:

1) Target velocity vector. This can be obtained from ship radar. If the jets were vectored to this target we can be nearly certain such data existed for ATIPs analysis.

2) Wind velocity vector at target altitude. This can obtained from the F18s. Every small piston engine aircraft I've ever flown in has a Garmin 1000 or similar GPS device capable of measuring ground velocity, air velocity, and subsequently calculating wind velocity. So there will be a complete trace of wind velocity of the jet at every altitude during its entire flight. It passed through the target's altitude during the flight, so they will have tight constraints on this vector.

With the above data, ruling out a balloon is an extremely simple matter: If the two vectors are not identical, it's not a balloon. End of story. ATIP claims to have ruled out balloon, and given these vectors are most certainly available to anyone in the Navy doing any analysis, I'm satisfied with that explanation. For that to be wrong, our intelligence analysts would have to be incapable of comprehending velocity vectors. Highly doubtful.

With the above data, ruling out a balloon is an extremely simple matter: If the two vectors are not identical, it's not a balloon. End of story. ATIP claims to have ruled out balloon, and given these vectors are most certainly available to anyone in the Navy doing any analysis, I'm satisfied with that explanation. For that to be wrong, our intelligence analysts would have to be incapable of comprehending velocity vectors. Highly doubtful.

ATIP = Advanced Aerospace Threat Identification Program (AATIP)?

We have to clarify. Is this the conclusion of the government agency or of the ex-AATIP employee Luis Elizondo?

To cite Luis Elizondo's opinion as unassailable is argument by authority.

Trusting the analysis of those who seemingly could not work out the height of the object even when they had 2 figures requiring pure simple conversion and maths to get a ball park of it.

ATIP = Advanced Aerospace Threat Identification Program (AATIP)?

We have to clarify. Is this the conclusion of the government agency or of the ex-AATIP employee Luis Elizondo?

To cite Luis Elizondo's opinion as unassailable is argument by authority.

Yes, the government intelligence group in charge of analyzing the video and everything else we're missing. What's the problem, exactly?

If it wasn't AATIP, the same goes for whichever government group with full data access did the analysis. They did not categorize this as a balloon. Again, all they would need is a simple radar track and wind velocity data to rule "balloon" out completely.

Last edited:
Trusting the analysis of those who seemingly could not work out the height of the object even when they had 2 figures requiring pure simple conversion and maths to get a ball park of it.

What's the correct target altitude? How would anyone determine that from FLIR if the range is unknown? All the Navy needs is a radar track to get the altitude. They don't even need the FLIR video. Same goes for target and wind velocity. Compare the two vectors. If they're not the same, you can scratch balloon off the go-fast list without knowing anything about the history of balloons or the history of pilot visual misidentification of said balloons.

Two vectors:

Case 1:
vector1.x = 34
vector1.z =124

vector2.x = 58
vector2.z = 81

Conclusion: They're not the same vector, therefore not a balloon.

Case 2:
vector1.x = 34
vector1.z =124

vector2.x = 34
vector2.z = 124

Conclusion: Same vector. Possible/probable balloon.

That's all there is to it.

Last edited:
Yes, the government intelligence group in charge of analyzing the video and everything else we're missing. What's the problem, exactly?

If it wasn't AATIP, the same goes for whichever government group with full data access did the analysis. They did not categorize this as a balloon. Again, all they would need is a simple radar track and wind velocity data to rule "balloon" out completely.
Are you sure this was anyone's conclusion? You seem rather unsure.

Still would be argument by authority.

Are you sure this was anyone's conclusion? You seem rather unsure.

Still would be argument by authority.

A program director reporting the results of his analysts is not an argument from authority. You're misusing the term.

Yes, he confirmed it vehemently with Mick in their conversation. Video is on YouTube. I am not unsure of that. What I'm unsure of his who's analysis resulted in the "unknown" status of this video (AATIP, Navy, or some other group(s) with full data access, there might have been more than one), not that AATIP concluded it was not a balloon. They did. The only way to conclude otherwise is to call Luis a liar or his analysts incapable of comparing "15" with "15" to see if they are the same number in a case where they have ship radar data.

Last edited:

Replies
54
Views
7K
Replies
48
Views
7K
Replies
24
Views
6K
Replies
15
Views
2K
Replies
5
Views
2K